Council Regulation (EC) No 2223/96 of 25 June 1996 on the European system of national and regional accounts in the Community

Including acquisitions less disposals of valuables.

Including acquisitions less disposals of non-financial assets.

Table 8.21 — Example of a more detailed social accounting matrix

Table 8.22 — Example of a detailed submatrix: net value added (basic prices)

CHAPTER 9

INPUT-OUTPUT FRAMEWORK

• 9.01. The input-output framework consists of three types of tables:

• (a) supply and use tables;

• (b) tables linking the supply and use tables to the sector accounts;

• (c) symmetric input-output tables.

• 9.02. Supply and use tables are matrices by industry and product describing the domestic production processes and the transactions in products of the national economy in great detail. These tables show:

• (a) the structure of the costs of production and the income generated in the production process;

• (b) the flows of goods and services produced within the national economy;

• (c) the flows of goods and services with the rest of the world.

• 9.03. A supply table shows the supply of goods and services by product and by type of supplier, distinguishing output by domestic industries and imports. A simplified example of a supply table is given in table 9.1.

Table 9.1 — A simplified supply table

• 9.04. A use table shows the use of goods and services by product and by type of use, i.e. as intermediate consumption (by industry), final consumption, gross capital formation or exports. Furthermore, the table shows the components of gross value added, i.e. compensation of employees, other taxes less subsidies on production, net mixed income, net operating surplus and consumption of fixed capital. A simplified example of a use table is given in table 9.2.

Table 9.2 — A simplified use table

• 9.05. Between the supply and use tables, two types of identities hold good (provided supplies and uses are valued consistently, see tables 9.5 and 9.6):

• (a) the identity by industry: Output by industry = Input by industry. In terms of our simplified supply and use table, this indicates that the row vector in cell (2,1) in table 9.1 should be equal to that in cell (3,1) in table 9.2. So, for each industry: Output = Intermediate consumption + Value added;

• (b) the identity by product: Total supply by product = Total use by product. In terms of our simplified tables, the column vector in cell (1,3) of table 9.1 should be equal to the column vector in cell (1,5) of table 9.2. So, for each product: Output + Imports = Intermediate consumption + Exports + Final consumption expenditure + Gross capital formation.

These identities by industry and product can be used to check and improve the consistency and completeness of estimates (see paragraph 9.11).

• 9.06. Supply and use tables are the central framework for all tables by industry, e.g. those on employment, gross fixed capital formation and capital stock.

• 9.07. The supply and use tables contain all the flows in the following accounts:

• (a) the goods and services account;

• (b) the production account;

• (c) the generation of income account.

• 9.08. A supply table and a use table can also be combined and presented as a single table. This can be achieved by adding two rows and a column to the table, for output and imports (see table 9.3). Note that the rows and columns from the supply table presented in paragraph 9.03 have been transposed here.

Table 9.3 — A simplified combined supply and use table

• 9.09. A symmetric input-output table is a product by product or industry by industry matrix describing the domestic production processes and the transactions in products of the national economy in great detail. A symmetric input-output table rearranges both supply and use in a single table. There is one major conceptual difference between a symmetric input-output table and a combined supply and use table: in the supply and use table, the statistics relate products to industries, while in the symmetric input-output table the statistics relate products to products or industries to industries. So, in a symmetric input-output table either a product or an industry classification is employed for both rows and columns (see table 9.4).

Table 9.4 — A simplified symmetric input-output table (product by product)

• 9.10. Most statistical information that can be obtained from producer units indicates what type of products they have produced/sold and, usually less detailed, what type of products they have bought/used. The format of the supply and use tables is designed to fit in with this type of statistical information (i.e. industry by product). By contrast, information of a product by product or industry by industry nature as required by the symmetric input-output table is not often available. For example, surveys of industries usually provide information about the type of products used and about the products produced. However, information on the inputs in terms of products and value added components for each product produced is usually not collectable. Ideally, the administration of an enterprise should show all costs allocated to the various types of output and, simultaneously, show the composition of intermediate consumption by type of product. In practice, information arranged in the form of supply and use tables is therefore a practical starting point for constructing the more analytic information in the symmetric input-output tables. The industry by product information in the supply and use tables can be converted into product by product or industry by industry statistics by adding extra statistical information on the input structures, or by assuming constant input structures by product or by industry (see paragraphs 9.54—9.60).

• 9.11. The supply and use tables serve both statistical and analytical purposes.

Important statistical purposes are:

• (a) identifying gaps and inconsistencies in basic data sources;

• (b) weighting and calculation of index numbers and price and volume measures;

• (c) making estimates by residual (estimating a variable by first estimating all other variables in the identity), e.g. for the production or final consumption of specific products;

• (d) checking and improving the consistency, plausibility and completeness of figures in the supply and use tables and the derived figures (such as those in the production accounts). To this end, the balancing process should not be limited to the supply and use tables at current prices:
• 1. by compiling supply and use tables at current and constant prices for two or more years, estimates of changes in volumes, values and prices can be balanced simultaneously: compared to integrating supply and use tables for a single year in isolation, this is a major extension of the effectiveness of the integration framework;
• 1. with the aid of the tables showing the linkage with the sector accounts, a direct comparison can be made with information from the latter, e.g. information on the distribution of income, on saving and on net lending (calculated as the result of financial transactions). This at least guarantees that, after the balancing process, consistency is obtained between the supply and use tables and the sector accounts;
• 1. trying to derive symmetric input-output tables from the supply and use tables may reveal inconsistencies and weaknesses in the supply and use tables. In this respect, there is therefore also a feedback from the symmetric input-output tables to the supply and use tables;

• (e) estimating figures for periods on which less reliable information is available, e.g. estimating annual figures on the basis of the detailed supply and use figures for a benchmark year or estimating quarterly figures on the basis of annual supply and use tables.

• 9.12. The supply and use tables and symmetric input-output tables give a detailed picture of the composition of the supply and use of goods, services and labour and the primary incomes involved. These tables and the ratios that can be derived from them, such as productivity figures, are an important subject for economic analysis.

• 9.13. The supply and use tables and symmetric input-output tables can also be used as tools of economic analysis. Both types of tables have different merits. For calculating direct and indirect effects, the supply and use tables need to be accommodated with specific assumptions or extra statistical information. For calculating cumulative effects, these assumptions and extra data requirements are the strongest. In fact, the requirements for calculating cumulative effects with a supply and use table amount to constructing a symmetric input-output table. Therefore, for calculating cumulative effects, the symmetric input-output table is the preferable tool. However, for calculating direct effects and first-order effects, the supply and use tables adjusted with a selected amount of assumptions (or extra statistical information) is in general to be preferred, because:

• (a) the calculation is less dependent on assumptions;

• (b) the supply and use table provides more detail than the symmetric input-output table;

• (c) the information in the supply and use table can be better linked to other types of statistical data.

These features are also helpful when the supply and use tables are integrated in a macro-economic model: the resulting overall model is closer to real statistics, can show a lot of detail and can relatively easily be linked to areas on which other statistical data are available, e.g. on the labour market or the environment.

• 9.14. The supply and use tables and symmetric input-output tables can be used to calculate:

• (a) effects of changes in prices or tax rates on the values of supply or use;

• (b) effects of changes in volumes on the values of supply or use;

• (c) effects of changes in prices of supply on prices of use;

• (d) effects of changes in the volume of use on the volume of supply;

• (e) effects of changes in the volume of supply on the volume of use.

The calculations can show indirect as well as direct effects. For example, a significant increase in energy prices will affect not only those industries that use energy intensively, but also those industries that use the outputs of the energy-intensive producers. With the aid of some assumptions, estimates of the size of such indirect effects can be deduced from the supply and use and symmetric input-output tables. Examples of common assumptions are:

• (a) a constant input structure in terms of values;

• (b) a constant composition of the value of output by industry and by product;

• (c) a constant composition of the value of final consumption expenditure of households by product.

These assumptions are rather rigid as they imply that relative prices do not change, that the production processes remain technically the same and that no substitution occurs between categories of final consumption expenditure by households. However, these general assumptions can be modified by allowing first for changes in relative prices, e.g. the Leontief-price model. This can then be extended with econometric or other estimates of the influence of relative prices and other variables on technical coefficients or final consumption expenditure by households.

The calculations need to be confined to the supply and use of goods and services. They could also be applied to the supply and use of labour and the components of value added.

• 9.15. The supply and use tables and the symmetric input-output table can be integrated into macro-economic models to provide the latter with a detailed meso-economic foundation. Specific types of analysis served by supply and use tables and the symmetric input-output table are, for example:

• (a) analysis of production, cost structures and productivity

• (b) analysis of prices;

• (c) analysis of employment;

• (d) analysis of the structure of capital formation, final consumption, exports, etc.;

• (e) analysis of the relationship between domestic production and the environment (e.g. focusing on the use of specific products like fuel, paper and glass);

• (f) analysis of imports of energy required;

• (g) impact analysis of new technologies;

• (h) sensitivity analysis of the effects of changes in tax rates and regulation.

SUPPLY AND USE TABLES

• 9.16. Tables 9.5 and 9.6 show the supply and use tables in more detail.

• 9.17. The classification used for industries in the  NACE Rev. 2 and the classification employed for products is the CPA; these classifications are fully aligned to each other: at each level of aggregation, the CPA shows the principal products of the industries according to the  NACE Rev. 2.

• 9.18. In the supply and use tables, the classification for products is at least as detailed as the classification for industries, e.g. the three digit-level of the CPA and the two digit-level of the  NACE Rev. 2.

• 9.19. The distinction between market output, output for own final use and other non-market output is only to be used for the total output by industry; the distinction is not required for each product group.

• 9.20. The distinction between market producers and producers for own final use on the one hand and other non-market producers on the other hand should only be used by industry when both types of producers are present within one industry. In general, this distinction will therefore only be used for subclassifying a very limited number of industries, e.g. health care and education (see paragraph 3.66.).

• 9.21. Imports and exports should be subdivided into:

• (a) intra-European Union deliveries;

• (b) imports and exports with other countries.

Table 9.5 — A supply table at basic prices, including a transformation into purchasers' prices

Table 9.6 — A use table at purchasers' prices

• 9.22. In the supply table, flows of goods and services are valued at basic prices. In the use table, the flows of goods and services are valued at purchasers' prices. In order to attain identities between supply and use, table 9.5 also shows the transition of supply at basic prices to supply at purchasers' prices. As supply should be equal to use, the table reveals also the transition of uses at purchasers' prices to uses at basic prices. As a consequence, two identities can be derived from this transition:

• (a) supply at purchasers' prices is equal to use at purchasers' prices;

• (b) supply at basic prices is equal to use at basic prices.

• 9.23. Value added is recorded at basic prices. It is the net result of output valued at basic prices less intermediate consumption valued at purchasers' prices.

• 9.24. Value added at factor cost is not a concept in the ESA. Nevertheless, it could be derived from value added at basic prices by subtracting other taxes less subsidies on production.

• 9.25. GDP is valued at market prices. This aggregate can be derived from the supply and use tables in three different ways:

• (a) the aggregate of output at basic prices by industry minus the aggregate of intermediate consumption at purchasers' prices by industry plus net taxes on products (the production approach);the intermediate consumption by industry includes the use of financial intermediation services indirectly measured that are recorded in a nominal industry (see paragraph 9.33);

• (b) the aggregates of the various components of value added at basic prices by industry minus the use of financial intermediation services indirectly measured (recorded in a nominal industry, see paragraph 9.33) plus the net taxes on products (the income approach);

• (c) the sum of final uses categories minus imports: exports - imports + final consumption expenditure + gross capital formation (all at purchasers' prices) (the expenditure approach).

• 9.26. The use table 9.6 also contains some supplementary information: gross fixed capital formation, stocks of fixed asssets and labour inputs by industry. This information is crucial for productivity analysis and may also serve several other types of analyses, e.g. analysis of employment.

• 9.27. In the ESA, the stocks of fixed assets are valued at the market prices prevailing on the data of the balance sheet. For fixed assets bought in earlier years, this implies that consumption of fixed capital of the foregoing years is to be deducted from the current market prices of new assets of the same type and quality. This net concept of fixed capital stock could be used in calculating capital intensity. However, for productivity analysis more often a specific gross concept of fixed capital stock is used. According to this gross concept all fixed assets should be valued at the current market prices of new assets of the same type and quality; no deduction should be made for capital consumption in the last and earlier years. This gross concept of fixed capital stock is not a standard concept in the ESA, but can easily be calculated with the aid of the Perpetual Inventory Method. Considering the great use of such gross figures, it is recommended that both gross and net figures on fixed capital stock be included as supplementary information.

• 9.28. For productivity analysis, hours worked is the preferable labour input variable. In drawing inferences about employed persons, however, the number of jobs may also be relevant. Both variables can be subclassified, e.g. into the part pertaining to employees and to self-employed persons.

• 9.29. For a proper compilation and understanding of the supply and use tables, it is important to recall some of the accounting conventions employed in the ESA:

• (a) the outputs of ancillary activities are not recorded separately; all the inputs consumed by an ancillary activity are treated as inputs into the principal or secondary activity which it supports. Ancillary activities may be, for example, marketing, accounting, storage and cleaning (see paragraphs 3.12 and 3.13);

• (b) goods or services produced and consumed within the same accounting period and within the same local KAU are not separately identified. They are therefore not recorded as part of the output or intermediate consumption of that local KAU. This may pertain for example to:
• 1. seeds and plans for sowing and planting;
• 1. hard coal consumed in the production of briquettes by coal mines;
• 1. electrical energy consumed by power stations;

• (c) minor processing, maintenance, servicing or repair on behalf of other local KAUs is to be recorded net, i.e. excluding the value of the goods involved. By contrast, when the goods are subject to a substantial physical change, the transaction should be recorded gross, i.e. as the purchase and sale of the goods involved;

• (d) durable goods can be rented or be subject to operating leasing. In such instances, they are recorded as fixed capital formation and fixed capital stock in their owner's industry; in the industry of the user intermediate consumption by amount of the rent is recorded;

• (e) persons working via temporary agencies are recorded as being employed in the industry of these agencies and not the industries in which they are actually working. As a consequence, in the latter industries, the payments for these persons are recorded as intermediate consumption and not as compensation of employees. Labour contracted out is treated similarly;

• (f) employment and compensation of employees in the ESA are broad concepts:
• 1. employment for social reasons is also counted as employment; this applies, for example, to work placements for disabled people, employment projects for people who have been unemployed for a long time and employment programmes for young people seeking jobs. As a consequence, the people involved are employees and receive compensation of employees (not social transfers), though their productivity may be (much) lower than that of the other employees;
• 1. employment includes also some cases in which the persons involved are not expected to work at all, e.g. persons dismissed but receiving for some period payments from their former employer. However, labour input in terms of hours worked is not distorted by this convention, as no hours are actually worked.

• 9.30. In the supply and use tables, two adjustment items should be introduced for reconciling the valuation of imports in the supply and use tables and in the sector accounts (see also tables 9.5 and 9.6).

In the supply table, imports of goods are valued at cif values. This is an overstatement of the costs of imports when the transport and insurance services incorporated in the cif value are provided by residents (e.g. own-account transport or transport by specialized resident carriers). In order to obtain the right balance between imports and exports, exports of services should thus also be increased by amount of this overstatement of imports. In the sector accounts, imports of goods are valued at fob values. This is an overstatement of imports as in the case of cif valuation. However, in the case of fob valuation the overstatement is smaller, which implies that the imputed increase in exports is smaller too. The results of employing different valuation principles is thus that net total imports are the same, but that both total imports and total exports are larger for cif valuation. The two valuation principles can be reconciled in the supply and use tables by introducing adjustment items for imports as well as exports. The adjustment items should be equal to the value of the transport and insurance services by residents incorporated in the cif value but not in the fob value, i.e. referring to the transport and insurance from the border of the exporting country to the border of the importing country. These adjustment items, once incorporated in the supply and use tables, need no special treatment in the input-output calculations.

• 9.31. The transfer of existing goods is recorded in the use table as a negative expenditure for the seller and a positive expenditure for the purchaser. For the product group involved, the transfer of an existing good amounts to a reclassification among uses. Only the transaction costs are not a reclassification: they are recorded as a use of business or professional services. For the purposes of description and analysis, it can be useful to show for some product groups the relative size of the transfer of existing goods separately, e.g. the importance of second-hand cars or the importance of recycled paper.

• 9.32. Direct purchases abroad by residents and purchases on the domestic territory by non-residents are commonly estimated as a special exercise in the statistical compilation process. They are introduced as adjustments to initial estimates of imports, exports and, by amount of the consumption part of the purchases abroad, final consumption expenditure. In order to obtain a balance between supply and use by product, all these purchases should be split over the various product groups involved. For the product groups where these types of purchases are important, they could be shown as a subcategory, e.g. expenditure on accommodation.

• 9.33. Throughout the supply and use tables, the  NACE Rev. 2 industry classification is extended with a nominal industry for the use of financial intermediation services indirectly measured. In the supply table, no transactions are recorded at all for this industry. In the use table, the total use of financial intermediation services indirectly measured is recorded as the intermediate consumption of this nominal industry. As this nominal industry does not have any other transactions, its net operating surplus is negative by the amount of its intermediate consumption; all other components of its value added are zero. As a consequence, its total gross value added is equal to its (negative) net operating surplus.

• 9.34. The transition from supply and use at basic prices to purchasers' prices involves:

• (a) reallocating trade margins;

• (b) reallocating transport margins;

• (c) adding taxes on products (except deductible VAT);

• (d) deducting subsidies on products.

The estimation of this transition is an important part of the balancing process. Separate tables can show the transition in more detail (see tables 9.7 and 9.8). These tables can also serve important analytical purposes, e.g. analysis of prices and analysis of the consequences of changes in the rates of taxes on products.

• 9.35. In general, when preparing supply and use tables and making the proper balancing between the two sides, there is always a choice of emphasis between two opposite ways of adjusting statistical data:

• (a) supply of each product at basic prices could be adjusted to a purchasers' price valuation to allow balancing with uses at purchasers' prices;

• (b) each of the uses at purchasers' prices could be adjusted to a basic prices' valuation to match with supply at basic prices.

• 9.36. In practice, both types of balances may be needed to build up a supply and use table. Both alternatives deal with or require similar kinds of adjustments, i.e. for taxes less subsidies on products and trade and transport margins by products. In fact, the first alternative is not possible without the second, since it is usually not possible to know the columns of taxes on products, subsidies on products and trade and transport margins broken down by products in the supply table unless the distribution among uses of the individual products is known from the use table at purchaser's prices (table 9.6).

• 9.37. Thus, the following tables are involved in the balancing process:

• (a) the supply and use tables 9.5 and 9.6 showing the final results of balancing totals of supply and use by products at purchasers' prices and at basic prices;

• (b) the tables on trade and transport margins and on taxes less subsidies on products (tables 9.7 and 9.8).

• 9.38. The transition of supply and use from basic prices to purchasers' prices (see paragraph 9.34) amounts to reallocating trade margins: valuation at basic prices implies that the trade margins are recorded as part of the product trade, while valuation at purchasers' prices implies that the trade margins are allocated to the products to which they pertain. An analogous situation holds for transport margins.

• 9.39. The total of trade margins by product is equal to the total of trade margins by the trade industries plus the secondary trade margins by other industries. An analogous equation holds for the transport margins.

Table 9.7 — A simple trade and transport margins table